AEI-Brookings Joint Center Policy Matters 05-19


Internet Backbone Competition and the SBC/AT&T Merger. Marius Schwartz. July 22, 2005.


“Internet backbone providers” operate transmission networks that serve as the arteries for Internet traffic.  These providers compete with one another for customers, such as retail Internet service providers, but must interconnect so each can provide its customers access to any Internet destination.  Opponents of the SBC/AT&T merger have argued that it threatens Internet backbone competition.  Given the stakes, this issue merits consideration.  But let’s consider the facts.

Five years ago, US and European competition authorities blocked the merger of MCI and Sprint for fear their combined backbone would be have such a dominant share of Internet traffic and users that it could force most or all rivals to start paying it for interconnection with its large user base.  Those concerns simply do not apply to the present merger.

At the time, MCI and Sprint were indisputably the top backbones, with collectively more than 50% of all US Internet traffic.  Today, AT&T has only about 12%, and faces strong rivals such as Level 3, Global Crossing, NTT/Verio, Qwest, and Sprint, to name but a few.  All of these carriers qualify to interconnect with AT&T at no charge, while SBC – though a major regional phone company – does not.  Moreover, SBC would add only about 7% to AT&T’s traffic share.

Critics have claimed that the merger should be analyzed jointly with that of Verizon/MCI and that the two merged entities should be presumed to act as a colluding pair, interconnecting with each other at no charge but making competitors pay.  Naturally, any firm would prefer to be paid if it could enforce such a demand.  But does the grim scenario hold water?  It does not, even setting aside the assertion of potential collusion.  Verizon’s backbone operations are smaller than SBC’s and MCI is a shadow of its former self, hence a merged Verizon/MCI would have a traffic share of less than 10%.  Thus SBC/AT&T and Verizon/ MCI together would have less than 30% of Internet traffic.  To put this in context, at the time of the Sprint merger, MCI by itself had a larger share – but still was unable to impose interconnection charges on fully eleven competitors! 

Critics have also singled out Verizon and SBC as large Internet service providers (ISPs) to residential and small business customers that use DSL for broadband Internet access, and stressed they could guarantee this traffic to their backbone merger partners.  However, most of this traffic is already reflected in the above shares.  Moreover, the spotlight on large broadband ISPs brings out a key point: nationally, Verizon and SBC together provide less than 28% of residential and small business broadband lines.  The rest is concentrated in the hands of cable giants Comcast and Time Warner, and a handful of other large broadband ISPs, including Cox, BellSouth, Charter, Cablevision, and Qwest.  These large ISPs have a keen interest in maintaining competitive backbone prices and can prevent any backbone from becoming dominant by steering their considerable traffic to other backbones.  Comcast alone has 7.4 million broadband lines, versus 5.6 for SBC and 3.9 for Verizon.

Sound competition policy in a market economy permits private transactions unless there is a good reason to oppose them.  Despite considerable scrutiny, no such reason has been presented for the Internet backbone.  The reviewing expert agencies should continue focusing on the facts, and reach the clear conclusion: the pending mergers pose no threat to Internet interconnection or prices.


The author is Professor of Economics at Georgetown University and has served as acting chief economist at the Antitrust Division of the U.S. Department of Justice and as senior economist at the President’s Council of Economic Advisers during the passage of the 1996 Telecom Act.  He has consulted for SBC on the Internet aspects of its merger.